The traditional colonoscopy has been shown to be a highly effective, lifesaving screening tool that can significantly reduce a person’s risk of dying from colon cancer. However, some abnormal growths that become cancer are still missed, and research indicates AI could help detect more of those and save patients money in the process.
One study in particular published in The Lancet Digital Health last month found that using Medtronic’s AI-powered system called GI Genius with a colonoscopy increased colorectal cancer prevention by nearly 5%. The Dublin, Ireland-based device maker’s AI system reduced incidence of colorectal cancer by 48.9, compared to no screening at all; while colonoscopy alone reduced the incidence of colorectal cancer by 44.2%.
The study—which was co-authored by a number of researchers who do consulting work for Medtronic—projected the reduction of colorectal cancer incidence would deliver a cost savings equal to $57 per patient. “At the U.S. population level, the implementation of AI detection during screening colonoscopy resulted in yearly additional prevention of 7,194 colorectal cancer cases and 2,089 related deaths, and a yearly saving of $290 million,” the researchers noted.
Medtronic’s GI Genius is the first system cleared by the FDA that uses AI to detect potential signs of colon cancer. The technology doesn’t replace the traditional colonoscopy but instead acts as a supplement to it.
A traditional colonoscopy is performed with a long, flexible tube called a colonoscope and a tiny affixed video camera that’s used by a physician to look for changes in the large intestine and rectum. That includes checking for small cell growths on the intestinal walls called polyps. Most polyps are benign, but some can eventually become cancerous. Once detected, these are typically removed easily. But some are missed.
“GI Genius detects colorectal polyps of all shapes and sizes automatically in real-time, which helps diagnose and potentially prevent colorectal cancer,” said Dr. Austin Chiang, chief medical officer for Medtronic’s gastrointestinal business. “The module serves as a second set of eyes that doesn’t get tired or distracted. By augmenting a doctor’s ability to detect polyps, AI has been shown to improve a physicians polyp detection rate during colonoscopy, reducing miss rates by half.”
At present, a number of academic medical institutions from MD Anderson in Houston to Rush University in Chicago are trying GI Genius out. But more study is still needed to determine the extent to which it might improve colon cancer detection and survival.
Cost and insurance coverage, too, are in flux. The projected savings from using GI Genius come, in part, from not needing to do as many screenings and catching cancer earlier or preventing it to lower healthcare costs. But Medtronic declined to disclose the cost of the technology to the health provider or patient.
At present, GI Genius remains the only AI-powered system for colonoscopy. But it’s part of a growing movement to improve all aspects of healthcare with artificial intelligence. Another company, EarlySign, for example, has a tool called ColonFlag that uses AI and machine learning to help identify patients with a high probability of having colorectal cancer.
“This is a really transformative technology,” Chiang said. “I think that in the future we’ll see artificial intelligence be a really important part of what we do for both diagnosis and treatment.”